Rank What to Do Next

Rank for what you should do next, not what would look impressive in a strategy deck.

Frequency, value, goals, feasibility, runnable now.

Feasibility includes effort, skill, data access, and change cost. Runnable in Copilot Chat or the Microsoft 365 app you already have should beat “needs a new platform.”

Each row needs a smallest proof: the first experiment, including a starter prompt or the next click.

Core principles

  1. Top five only. Everything else stays on Not now.
  2. High frequency plus high pain plus high feasibility rises.
  3. Name the Copilot surface on each row.
  4. Under the table: what Copilot inferred, and what context it still wished it had.

Check yourself

What should a top-five row include besides the idea?

Do this in Copilot

Leave with a top-five table you would show a manager without embarrassment.

Save the table. Do not run the experiment until Lesson 5 unless you are ahead and have time.

Build the ranked table

Using only these candidates: [PASTE FILTERED CANDIDATES]. Produce one prioritized table of my top 5 Copilot use cases. Columns: Rank; Use case (one sentence, specific to my world); Frequency (H/M/L); Value per time (H/M/L); Impact on my stated goals (H/M/L); Feasibility (H/M/L); Runnable in Copilot now? (Yes / Partial / Needs something else) — name the app; First experiment — smallest proof, including a starter prompt. Then 2–3 lines on what you inferred, and 1 line on context you still wish you had. Then stop and ask whether I want to go deeper on one row or stress-test feasibility on one row. Wait.

Techniques: Role + task + format · Constraints

Open Copilot

Did you run this in Copilot? Mark complete when you have tried it.

Next lesson: Run the First Experiment →

AI

Part of the MillerKnoll collective

Studio · 2026